2022 International Conference on Quality Management, Transport and Information Security, Information Technologies (IT&QM&IS),
Год журнала:
2023,
Номер
unknown, С. 133 - 136
Опубликована: Сен. 25, 2023
Diseases
of
the
cardiovascular
system
are
main
cause
death
in
world
population.
Classification
electrocardiogram
(ECG)
signals
is
a
reliable
method
for
diagnosing
cardiac
pathologies.
The
available
ECG
databases
consist
an
unequal
number
from
various
This
article
analyzes
impact
using
class
alignment
methods
on
result
neural
network
classification
signals.
results
demonstrate
that
SMOTE
GRU
algorithm
provides
high
performance
classifying
segments,
while
BiLSTM
ROS
full
Accuracy,
Loss,
Recall,
Precision,
F-score
values
respectively
70.31%
and
77.73%,
0.29
0.41,
90.1%
96.0%,
78.8%
83.4%,
88.5%
95.3%.
Increased
focus
is
being
positioned
on
the
potential
of
"smart
grid"
idea
to
improve
efficiency
power
production
and
delivery.
Reduce
energy
usage
make
use
better
resources.
The
study
smart
grids
starting
point
for
a
wide
range
related
fields.
Energy
use,
waste
reduction,
database
optimization,
an
effective
communication
system
are
all
in
this
category.
goal
chapter
suggest
architecture
making
most
renewable
sources.
suggested
collects
consumption
profile
heterogeneous
devices
using
Internet
things
principles.
microgrid
compiles
data
create
timetable
particular
gadgets,
which
it
then
broadcasts.
By
decreasing
need
expensive
outside
sources,
demonstrates
effectiveness
design.
implementation
detailed
many
robotics
cases.
Reducing
greenhouse
gas
emissions
requires
smart
buildings.
Growing
in
popularity,
machine
learning
(ML)
may
improve
decarbonization
management
and
analytics
for
It's
an
essential
tool
many
industries,
including
cities.
Grid
consumers
are
seeing
the
emergence
of
energy
communities.
Buildings
can
learn
thanks
to
artificial
intelligence
(AI)
ML.
Learn
about
capabilities
ML
algorithms
systems.
In
this
chapter,
we
shall
define
machine-based
provide
a
general
overview
smart-based
We
will
examine
these
algorithms'
advantages,
difficulties,
practical
uses.
also
go
over
important
implementation
concerns
future
developments
fascinating
sector.
Renewable
energy
and
the
economy
are
closely
intertwined,
with
renewable
sources
like
solar
wind
contributing
to
economic
growth
by
creating
jobs
reducing
costs.
The
widespread
adoption
of
can
lead
increased
security,
reduced
greenhouse
gas
emissions,
lower
bills,
all
which
have
positive
impacts
on
economy.
impact
is
multifaceted.
It
stimulates
through
job
creation
in
sector,
reduces
import
expenses,
enhances
resilience,
ultimately
leading
a
more
stable
sustainable
generation
involves
harnessing
natural
resources
sunlight,
wind,
water
order
generate
power,
turbines
panels,
hydropower
facilities
common
methods,
technological
advancements
continue
make
efficient
cost-effective,
further
bolstering
their
contribution
mix
Other
mechanisms
for
generating
include
geothermal
energy,
added
hot
from
world's
core,
tidal
utilizes
moon's
gravitational
pull-on
oceans.
Wave
biomass
come
organic
wood
agricultural
waste,
ocean
waves'
kinetic
also
contribute
mix.
However,
these
limitations.
Geothermal
location-dependent,
limiting
its
applicability.
Tidal
disrupt
marine
ecosystems
faces
challenges
converting
intermittent
tides
steady
power
source.
Biomass
energy's
carbon
neutrality
debated
due
emissions
processing
transportation.
devices
costly
susceptible
wear
tear
harsh
conditions,
while
life
needs
careful
consideration.
Controlling
and
monitoring
energy
use
requires
accurate
data,
this
is
what
management
systems
(EMS)
deliver.
Using
Internet
of
things
(IoT)-based
technology,
these
EMS
may
be
vastly
improved
upgraded,
resulting
in
more
savings.
This
research
provides
support
for
the
real-time
IoT
eco-friendly
smart
structures.
Taking
readings
use,
making
forecasts
future
recognizing
people's
faces
are
three
cornerstones
proposed
system.
Predictions
were
made
using
a
method
called
short-term
load
forecasting
(STLF)
that
based
on
K-nearest
neighbor
(KNN)
algorithm.
Line
A
current,
line
B
C
voltage
A,
volt
B,
six
digital
power
meter
(DPM)
parameters
must
utilized
as
data
to
serve
training
prediction
algorithms.
The
building's
usage
subsequent
hours
same
day
calculated
predicted
outcome.
Based
outcome,
active,
reactive,
seeming
abilities
determined.
facial
recognition
administrators
restrict
access
restricted
areas.
Viola-Johns
algorithm
foundation
modern
technology.
system
has
total
accuracy
91%
face
detection
recognition,
measured
by
classifier's
Haar
characteristics.
results
showed
true
negative
rate
(TNR),
positive
predictive
value
(PPV),
false
discovery
(FDR)
each
averaged
51%,
whereas
PPV
70.5%
FDR
31.6%.
International Journal of Electrical and Electronics Engineering,
Год журнала:
2024,
Номер
11(11), С. 326 - 340
Опубликована: Ноя. 30, 2024
An
essential
diagnostic
technique
for
assessing
cardiac
health
is
an
Electrocardiogram
(ECG).
The
heart's
electrical
activity
captured
in
this
recording.
need
to
share
the
workload
among
physicians
and
relieve
pressure
on
them
has
led
development
of
automatic
detection
classification
techniques
heart
arrhythmias
other
abnormalities
as
number
patients
increased.
All
operate
following
stages:
signal
preprocessing,
which
includes
denoising,
extracting
features,
categorising
features.
Recently,
several
methods
have
been
used
denoise,
extract
categorize
ECG
signals.
preprocessing
necessary
before
extraction
phase
because
numerous
noise
sources
a
medical
setting
can
deteriorate
signal.
present
study
reviews
analysis,
feature
extraction,
denoising
techniques.
Frequency
domain
filters
adaptive,
Wavelet
Transform
(WT)
based
are
commonly
denoise
For
ultimate
task,
various
morphological,
temporal,
statistical
Fourier
transform,
wavelet-based
coefficients
frequently
extracted
from
Findings
show
that
deep
learning
best
others
task
hybrid
features
increase
efficacy.
Most
authors
attempted
into
five
classes.
There
scope
identify
combine
most
effectively
provide
better
performance
more
diseases.
Also,
there
developing
classifier
performs
classify
significant
or
Journal of Intelligent & Fuzzy Systems,
Год журнала:
2023,
Номер
unknown, С. 1 - 16
Опубликована: Ноя. 22, 2023
Arrhythmia
disorders
are
the
leading
cause
of
death
worldwide
and
primarily
recognized
by
patient’s
irregular
cardiac
rhythms.
Wearable
Internet
Things
(IoT)
devices
can
reliably
measure
patients’
heart
rhythms
producing
electrocardiogram
(ECG)
signals.
Due
to
their
non-invasive
nature,
ECG
signals
have
been
frequently
employed
detect
arrhythmias.
The
manual
procedure,
however,
takes
a
long
time
is
prone
error.
Utilizing
deep
learning
models
for
early
automatic
identification
arrhythmias
preferable
approach
that
will
improve
diagnosis
therapy.
Though
analysis
using
cloud-based
methods
perform
satisfactorily,
they
still
suffer
from
security
issues.
It
essential
provide
secure
data
transmission
storage
IoT
medical
because
its
significant
development
in
healthcare
system.
So,
this
paper
proposes
arrhythmia
classification
system
with
help
effective
encryption
(DL)
proposed
method
mainly
involved
two
phases:
signal
disease
classification.
In
phase,
collected
through
sensors
encrypted
optimal
key-based
elgamal
elliptic
curve
cryptography
(OKEGECC)
mechanism,
securely
transmitted
cloud.
After
that,
collects
Massachusetts
Institute
Technology-Beth
Israel
Hospital
(MIT-BIH)
database
training.
preprocessed
applying
continuous
wavelet
transform
(CWT)
quality
data.
Next,
feature
extraction
carried
out
deformable
attention-centered
residual
network
50
(DARNet-50),
finally,
performed
butterfly-optimized
Bi-directional
short-term
memory
(BOBLSTM).
experimental
outcomes
showed
achieves
99.76%
accuracy,
which
better
than
existing
related
schemes.
2022 International Conference on Quality Management, Transport and Information Security, Information Technologies (IT&QM&IS),
Год журнала:
2023,
Номер
unknown, С. 133 - 136
Опубликована: Сен. 25, 2023
Diseases
of
the
cardiovascular
system
are
main
cause
death
in
world
population.
Classification
electrocardiogram
(ECG)
signals
is
a
reliable
method
for
diagnosing
cardiac
pathologies.
The
available
ECG
databases
consist
an
unequal
number
from
various
This
article
analyzes
impact
using
class
alignment
methods
on
result
neural
network
classification
signals.
results
demonstrate
that
SMOTE
GRU
algorithm
provides
high
performance
classifying
segments,
while
BiLSTM
ROS
full
Accuracy,
Loss,
Recall,
Precision,
F-score
values
respectively
70.31%
and
77.73%,
0.29
0.41,
90.1%
96.0%,
78.8%
83.4%,
88.5%
95.3%.
